@hybridlabor-api/aos 4.4.0 → 4.4.2-beta.0

package/mcps/davinci-resolve-mcp/.claude/skills/media-analysis.md

@@ -1,44 +0,0 @@
----
-name: resolve-media-analysis
-description: Media intelligence layer for the DaVinci Resolve MCP. Uses FFprobe, FFmpeg, and optionally Whisper to READ and ANALYZE source media — never modify, transcode, convert, or create derivatives. Provides the MCP with full context of what footage actually is so it can take informed actions within Resolve.
----
-
-# Resolve Media Analysis — Claude Code Skill
-
-This skill wraps the project's canonical media analysis guide with Claude Code-specific integration.
-
-**Read `docs/media-analysis-guide.md` for the complete guide.** Everything below is Claude Code-specific context.
-
-## The First Rule
-
-**Never touch the source.** Your relationship to source media is READ-ONLY. See `docs/media-analysis-guide.md` section "The First Rule: Never Touch the Source" for the full rationale from every post-production department.
-
-## MCP Integration
-
-When using the DaVinci Resolve MCP tools alongside media analysis:
-
-### Getting File Paths from Resolve
-
-1. **From media pool clips:** `media_pool_item` -> `get_clip_property(clip_id)` returns `"File Path"`
-2. **From timeline items:** `timeline_item` -> `get_media_pool_item(item_id)` -> then get clip properties
-3. **From media storage:** `media_storage` -> `get_files(path)` lists files in a directory
-
-### Workflow: Analyze Before Acting
-
-1. **Identify the media** — Use MCP to get clip IDs and file paths
-2. **Check for existing analysis** — Look for sidecar JSON files
-3. **Analyze if needed** — Run FFprobe (+ optional tools) on the files
-4. **Act with context** — Use MCP tools with full knowledge of what the media is
-
-Analysis informs Resolve actions. At no point do we create intermediate files that enter the media pipeline.
-
-## Setup
-
-On first use, ask the user the three setup questions documented in the guide:
-1. Where to save analysis files (alongside / directory / project)
-2. Tool detection (FFprobe required, FFmpeg/Whisper optional)
-3. Analysis depth (quick / standard / deep)
-
-## Full Reference
-
-All analysis commands, output format, examples, proactive warnings, and principles are in `docs/media-analysis-guide.md`. Do not duplicate that content — read and follow it.

package/mcps/davinci-resolve-mcp/.clinerules

@@ -1,17 +0,0 @@
-# DaVinci Resolve MCP Server — Cline Rules
-
-## Project
-Python MCP server (30 compound tools, 328 granular tools in `--full` mode) for controlling DaVinci Resolve via its Scripting API. See `AGENTS.md` for full project overview and the README for the 109-action kernel coverage map.
-
-## Critical Rule: Source Media Integrity
-NEVER modify, transcode, convert, or create derivatives of source media files unless the user explicitly requests it. All analysis is read-only. See `docs/media-analysis-guide.md` for the complete media analysis workflow and rationale.
-
-## Media Analysis
-When working with media files, use FFprobe/FFmpeg to analyze (read-only) before taking actions in Resolve. The full guide is at `docs/media-analysis-guide.md`.
-
-## Key Files
-- `src/server.py` — Main compound MCP server (30 tools)
-- `src/resolve_mcp_server.py` — Full server (328 tools)
-- `src/utils/` — Platform detection, Resolve connection, helpers
-- `docs/media-analysis-guide.md` — Media analysis workflow guide
-- `install.py` — Universal installer for 10+ MCP clients

package/mcps/davinci-resolve-mcp/.cursorrules

@@ -1,17 +0,0 @@
-# DaVinci Resolve MCP Server — Cursor Rules
-
-## Project
-Python MCP server (30 compound tools, 328 granular tools in `--full` mode) for controlling DaVinci Resolve via its Scripting API. See `AGENTS.md` for full project overview and the README for the 109-action kernel coverage map.
-
-## Critical Rule: Source Media Integrity
-NEVER modify, transcode, convert, or create derivatives of source media files unless the user explicitly requests it. All analysis is read-only. See `docs/media-analysis-guide.md` for the complete media analysis workflow and rationale.
-
-## Media Analysis
-When working with media files, use FFprobe/FFmpeg to analyze (read-only) before taking actions in Resolve. The full guide is at `docs/media-analysis-guide.md`.
-
-## Key Files
-- `src/server.py` — Main compound MCP server (30 tools)
-- `src/resolve_mcp_server.py` — Full server (328 tools)
-- `src/utils/` — Platform detection, Resolve connection, helpers
-- `docs/media-analysis-guide.md` — Media analysis workflow guide
-- `install.py` — Universal installer for 10+ MCP clients

package/mcps/davinci-resolve-mcp/.github/copilot-instructions.md

@@ -1,17 +0,0 @@
-# DaVinci Resolve MCP Server — GitHub Copilot Instructions
-
-## Project
-Python MCP server (30 compound tools, 328 granular tools in `--full` mode) for controlling DaVinci Resolve via its Scripting API. See `AGENTS.md` for full project overview and the README for the 109-action kernel coverage map.
-
-## Critical Rule: Source Media Integrity
-NEVER modify, transcode, convert, or create derivatives of source media files unless the user explicitly requests it. All analysis is read-only. See `docs/media-analysis-guide.md` for the complete media analysis workflow and rationale.
-
-## Media Analysis
-When working with media files, use FFprobe/FFmpeg to analyze (read-only) before taking actions in Resolve. The full guide is at `docs/media-analysis-guide.md`.
-
-## Key Files
-- `src/server.py` — Main compound MCP server (30 tools)
-- `src/resolve_mcp_server.py` — Full server (328 tools)
-- `src/utils/` — Platform detection, Resolve connection, helpers
-- `docs/media-analysis-guide.md` — Media analysis workflow guide
-- `install.py` — Universal installer for 10+ MCP clients

package/mcps/davinci-resolve-mcp/.github/workflows/npm-publish.yml

@@ -1,109 +0,0 @@
-name: Publish npm package
-
-on:
-  push:
-    tags:
-      - "v*"
-
-permissions:
-  contents: read
-  id-token: write
-
-jobs:
-  publish:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Check out repository
-        uses: actions/checkout@v5
-
-      - name: Set up Python
-        uses: actions/setup-python@v6
-        with:
-          python-version: "3.11"
-
-      - name: Set up Node
-        uses: actions/setup-node@v5
-        with:
-          node-version: "24"
-          registry-url: "https://registry.npmjs.org"
-
-      - name: Run Python smoke tests
-        run: python tests/test_import.py
-
-      - name: Run static guard tests
-        # Drift guards that gate every publish: no undefined names in src/,
-        # tool action lists in sync with dispatch, control-panel guide in
-        # sync with the panel's navigation and screenshots, and the
-        # Blackmagic-facing api-limitations.md in sync with the api_truth ledger.
-        run: |
-          pip install pyflakes
-          python -m unittest tests.test_static_undefined_names tests.test_action_list_drift tests.test_panel_docs_drift tests.test_api_limitations_doc
-
-      - name: Run Node CLI smoke tests
-        run: |
-          node bin/davinci-resolve-mcp.mjs --help
-          node bin/davinci-resolve-mcp.mjs --version
-
-      - name: Run setup dry run
-        run: |
-          export DAVINCI_RESOLVE_MCP_INSTALL_ROOT="$(mktemp -d)"
-          node bin/davinci-resolve-mcp.mjs setup --dry-run --no-venv --clients manual --update-policy never
-
-      - name: Check npm package contents
-        run: npm pack --dry-run
-
-      - name: Check npm version availability
-        id: npm_version
-        shell: bash
-        run: |
-          package_name="$(node -p "require('./package.json').name")"
-          package_version="$(node -p "require('./package.json').version")"
-          if npm view "${package_name}@${package_version}" version >/dev/null 2>&1; then
-            echo "publish=false" >> "$GITHUB_OUTPUT"
-            echo "${package_name}@${package_version} is already published; skipping npm publish."
-          else
-            echo "publish=true" >> "$GITHUB_OUTPUT"
-            echo "${package_name}@${package_version} is not published yet."
-          fi
-
-      - name: Compute expected tarball checksum
-        id: pack_meta
-        if: steps.npm_version.outputs.publish == 'true'
-        shell: bash
-        run: |
-          shasum="$(npm pack --dry-run --json | node -p "JSON.parse(require('fs').readFileSync(0, 'utf8'))[0].shasum")"
-          echo "shasum=${shasum}" >> "$GITHUB_OUTPUT"
-          echo "Expected tarball shasum: ${shasum}"
-
-      - name: Publish to npm
-        id: npm_publish
-        if: steps.npm_version.outputs.publish == 'true'
-        continue-on-error: true
-        run: npm publish --provenance --access public
-
-      - name: Verify publish landed
-        if: steps.npm_version.outputs.publish == 'true' && steps.npm_publish.outcome == 'failure'
-        shell: bash
-        run: |
-          # npm publish can exit nonzero after the registry has already accepted
-          # the PUT: a retried request re-presents the consumed single-use OIDC
-          # exchange token and gets 401 "token is invalid" (observed on the
-          # v2.37.0 run). Only fail the job if the exact tarball is not live.
-          package_name="$(node -p "require('./package.json').name")"
-          package_version="$(node -p "require('./package.json').version")"
-          expected="${{ steps.pack_meta.outputs.shasum }}"
-          for attempt in 1 2 3 4 5 6; do
-            published="$(npm view "${package_name}@${package_version}" dist.shasum 2>/dev/null || true)"
-            if [ -n "${published}" ]; then
-              if [ "${published}" = "${expected}" ]; then
-                echo "npm publish reported failure, but ${package_name}@${package_version} is live with the expected shasum ${expected}. Treating as success."
-                exit 0
-              fi
-              echo "::error::${package_name}@${package_version} exists on the registry with shasum ${published}, expected ${expected}. Refusing to mask the publish failure."
-              exit 1
-            fi
-            echo "Attempt ${attempt}: ${package_name}@${package_version} not visible on the registry yet; retrying in 10s."
-            sleep 10
-          done
-          echo "::error::npm publish failed and ${package_name}@${package_version} never appeared on the registry."
-          exit 1

package/mcps/davinci-resolve-mcp/.windsurfrules

@@ -1,17 +0,0 @@
-# DaVinci Resolve MCP Server — Windsurf Rules
-
-## Project
-Python MCP server (30 compound tools, 328 granular tools in `--full` mode) for controlling DaVinci Resolve via its Scripting API. See `AGENTS.md` for full project overview and the README for the 109-action kernel coverage map.
-
-## Critical Rule: Source Media Integrity
-NEVER modify, transcode, convert, or create derivatives of source media files unless the user explicitly requests it. All analysis is read-only. See `docs/media-analysis-guide.md` for the complete media analysis workflow and rationale.
-
-## Media Analysis
-When working with media files, use FFprobe/FFmpeg to analyze (read-only) before taking actions in Resolve. The full guide is at `docs/media-analysis-guide.md`.
-
-## Key Files
-- `src/server.py` — Main compound MCP server (30 tools)
-- `src/resolve_mcp_server.py` — Full server (328 tools)
-- `src/utils/` — Platform detection, Resolve connection, helpers
-- `docs/media-analysis-guide.md` — Media analysis workflow guide
-- `install.py` — Universal installer for 10+ MCP clients

package/mcps/davinci-resolve-mcp/AGENTS.md

@@ -1,114 +0,0 @@
-# DaVinci Resolve MCP Server — Agent Instructions
-
-This is the canonical short operating brief for AI coding agents working in this
-repository. Keep it durable: do not duplicate version numbers, tool counts,
-kernel-action counts, release notes, or long reference tables here.
-
-## Non-Negotiable Source Media Safety
-
-Never modify, transcode, convert, proxy, relink, replace, or create derivatives
-of source media unless the user explicitly asks for that exact operation.
-
-Analysis workflows may read source media, but file outputs must go to sidecar
-files, session scratch space, or the configured
-`davinci-resolve-mcp-analysis` project root. Resolve-target analysis should run
-visual analysis, transcription, metadata writeback, and Media Pool marker
-writeback by default unless the user opts out; those writes are Resolve project
-database changes, not source-media changes. Preserve the chain from camera
-original to final delivery.
-
-See `docs/guides/media-analysis-guide.md` for the full source-safe workflow.
-
-## Media Analysis Defaults Are Mandatory
-
-Do not translate source-safe into underpowered, read-only, no-writeback analysis.
-When the user asks to analyze Resolve media, run the requested analysis directly
-with visual analysis, transcription, persisted artifacts, metadata writeback, and
-Media Pool marker writeback enabled by default. Do not add
-`include_visuals=false`, `include_transcription=false`,
-`publish_metadata=false`, `timed_markers=no`, `session_only=true`, or
-`dry_run=true` unless the user explicitly asks for that opt-out, the target is a
-raw file path that cannot receive Resolve project writeback. Vision uses
-`host_chat_paths` by default — analyze actions return a deferred payload with
-absolute `frame_paths` and a JSON schema; the host chat must read those frames
-as local images, produce JSON per the schema, and call
-`media_analysis(action="commit_vision", params={clip_id, visual, vision_token})`
-to merge the result and trigger metadata + Media Pool clip-marker writeback.
-Not completing `commit_vision` leaves the analysis in
-`pending_host_vision_analysis` — that is a failure mode, not a success. Users
-must explicitly opt out with `include_visuals=false` for a technical-only run.
-
-## Frame-Referenced Color Work
-
-Before applying a grade, look, shot match, LUT, CDL, DRX, or copied grade to an
-existing Resolve timeline, inspect representative Resolve-rendered frames for
-the target shot or shots. Use thumbnails, contact sheets, Gallery stills, marker
-frames, or visual analysis reports written only to scratch/analysis locations.
-When the API can safely provide them, compare matched untreated/bypass, current,
-and after frames at the same timecodes; restore the previous active version or
-node-enabled state after any temporary bypass capture.
-
-Do not grade only from clip metadata, graph availability, or a requested style
-label unless the user explicitly asks for a blind/global pass. Preserve or create
-a recoverable grade version, and report which frame references informed the
-change. Do not describe Resolve's default one-node graph as an existing creative
-grade unless active grade tools, LUTs, or other grade state are present.
-
-When the user asks to build on or adjust an existing grade, treat the current
-grade as creative work to preserve. Inspect the active grade version and node
-graph first, create or switch to a recoverable adjustment version, and make
-incremental changes only through supported controls. Do not reset grades, replace
-graphs, or apply whole-grade artifacts unless the user explicitly asks for that
-semantics.
-
-## Source Of Truth
-
-- Public overview, current stats, and docs map: `README.md`
-- Historical release notes: `CHANGELOG.md`
-- AI assistant operating reference: `docs/SKILL.md`
-- Release checklist and validation rules: `docs/process/release-process.md`
-- Kernel workflow support maps: `docs/kernels/`
-- API coverage and live-test status: `docs/reference/api-coverage.md`
-- Blackmagic-facing API gaps/bugs (generated from `api_truth`):
-  `docs/reference/api-limitations.md` — when you document a new Resolve API
-  limitation, add a `submit`-tagged entry to `src/utils/api_truth.py` and
-  regenerate with `scripts/gen_api_limitations.py` (a drift guard enforces it)
-- Bundled Resolve API text: `docs/reference/resolve_scripting_api.txt`
-
-## Key Paths
-
-- Compound server: `src/server.py`
-- Granular server entrypoint: `src/resolve_mcp_server.py`
-- Local control panel launcher: `src/control_panel.py`
-- Granular implementation: `src/granular/`
-- Utilities: `src/utils/`
-- Installer: `install.py`
-- Tests: `tests/`
-- Examples: `examples/`
-
-## Common Commands
-
-```bash
-python src/server.py
-python src/server.py --full
-venv/bin/python -m src.control_panel
-python install.py
-venv/bin/python tests/test_import.py
-venv/bin/python scripts/audit_api_parity.py
-```
-
-Python 3.10+ is required (the MCP SDK floor). 3.10-3.12 is the lowest-risk range
-for Resolve scripting; 3.13/3.14 are accepted and verified on Resolve Studio
-20.3.2, but older Resolve builds may fail to connect on 3.13+.
-
-## Development Notes
-
-- Prefer the compound server unless a task specifically needs granular tools.
-- Use existing action-dispatch patterns and helper functions before adding new
-  abstractions.
-- All Resolve-facing temp or export paths should use the repo's safe path
-  helpers; do not invent ad hoc temp paths for files Resolve writes.
-- For changes touching Resolve behavior, update focused tests and follow the
-  live-validation guidance in `docs/process/release-process.md`.
-- For docs changes, keep the README concise and move durable detail into
-  dedicated files under `docs/`.